This renewal application is being submitted after the 1st year of funding under the R03 mechanism. The goal is to establish the value of genetic alterations in predicting the natural history and responsiveness to adjuvant chemotherapy of Dukes' B and C colon cancer utilizing patient material from cooperative group trials. In the 1st year of limited funding, we made significant progress in 2 areas: first, the experiments were formally adopted as Eastern Cooperative Oncology Group (ECOG) protocol EST 5290, establishing mechanisms for sample acquisition, and data recording, submission and statistical analysis by the ECOG statistical component at the Dana Farber; second,, we rigorously determined the level of c-myc copy number in the DNA from over 324 clinical samples (tumor+normal) encompassing 139 patients entered on ECOG protocol EST 2284, and began to collect samples from ECOG protocol EST 2288 (301 patients thus far accrued to our protocol from a total accrual to the parent protocol of 874 ECOG patients and 2592 in the Intergroup study; total projected accrual to ECOG/Intergroup approx. 900/2700, respectively, which will yield greater than 3,000 samples). The specific aims of this renewal application are similar to those originally funded as a small pilot project: to determine the prognostic value of amplifications of c-myc and structural alterations (mutations and deletions) at 17pl2 (the p53 gene) and 18q2l (the DCC gene) in both the natural history of Dukes' B and C colon tumors and in predicting outcome with adjuvant therapy. Analysis of DNA from tumors of a recently completed ECOG/Intergroup protocol (EST 2284) will be used to develop a statistical model which will be tested prospectively in a second, on-going, ECOG/Intergroup protocol (EST 2288) of adjuvant therapy.